Bidding Wind Power on the Day-Ahead Market by Minimizing the Imbalance Costs.
MetadataVis full innførsel
- Institutt for elkraftteknikk 
For wind power to participate in the short-term power market, forecasts up to 36 hours ahead of the hour of operation are required. These forecasts are never perfect, and the resulting wind power mismatch implies imbalance costs for the participant in the regulation market. Bidding the expected forecast in the day-ahead market is sub-optimal, as the difference between the day-ahead spot price and upward and downward regulation price is different. If the traded volume on the regulation market is low and the mismatched energy is high, this will have an impact on the imbalance price. This research aims to find a bidding strategy for stand-alone wind power in the day-ahead market that includes the price elasticity of the regulation market. The imbalance costs can be further reduced by teaming the wind power with conventional power that can balance out the mismatches. Hence, a bidding strategy for a mixed portfolio is also proposed. For both bidding strategies the value of the wind forecast service and its accuracy are also simulated. The ethics and social welfare of the bidding strategies are then examined. The bidding strategies were simulated in the Dutch market for 2006. For stand-alone wind power in the day-ahead market the proposed bidding strategy resulted in a profit increase of 3.5% compared to the well-known optimal quantile bidding strategy and 7.5% compared to bidding the expected generation. Teaming the wind power with gas power to balance out the mismatch gave a profit increase of 9.6% compared to separate bidding in the day-ahead market. The profit of teaming is greatly reduced by the capacity reserves that the conventional unit need to withhold in order to balance out the mismatch. In reality, the excess reserve capacity can be bid at the balancing market when the hour of operation approaches and the expected wind power generation can be more accurately estimated. This strategy may increase the mixed portfolio profit substantially. For wind farms bidding alone in the day-ahead market the profits are strongly affected by the level of wind forecast accuracy. For the 200MW capacity wind farm used in this simulation, the profit increase was 11 million euros for a 10 percentage point decrease in the standard deviation of the wind forecast error, if the initial forecast had 25% standard deviation. The dependency was greatly reduced when teamed with a 400MW conventional unit. The same decrease in wind forecast error now resulted in a profit increase of 5 million euros. The bidding strategies do not violate bidding rules and was found to decrease the total amount of regulation power needed. It is not possible to exploit the difference in market price between the day-ahead market and the regulation market, as assumptions the bidding strategy is based upon prevent the possibility of market arbitrage. The cooperation between the conventional unit and the wind farm will create a player in the market that has more power to affect the prices versus participants bidding separately in the market. However, the bidding strategy is based on a price-taking assumption for the day-ahead market and since any involvement in the balancing market is assumed to only decrease the profit, no market power will be exploited using this strategy. Social welfare is unchanged by the bidding strategy. Depending on the regulation state and the traded imbalance value, there is a transfer of surplus between supplier and consumer in the day-ahead market and the regulation market. For the wind farm there is a total gain of surplus, making wind power a more compatible investment option versus other conventional power sources to meet the rising electricity demand. As an environmental energy source this can have a positive impact on the social welfare in the long run. However, examining the environmental benefits is a more complex analysis including analysis of the environmental effects of the balance responsible conventional unit.